A Currency That Worked for 700 Years
Every stablecoin whitepaper opens the same way: a mechanism, a peg, a reserve, an attestation schedule. Here is a different opening. In the year 312, a mint in Constantinople started producing a gold coin at seventy-two to the Roman pound — about 4.5 grams — and then simply refused to change it for roughly seven hundred years.
No peg. No reserve. No oracle. The solidus was the reserve. Its stability was not a claim about assets held elsewhere; it was a property of the object itself, verifiable by anyone with a scale. Merchants from Scandinavia to Sri Lanka accepted it — hoards still surface on both ends of that range — because the coin tomorrow would be the coin today. Around the year 550, the Alexandrian merchant Cosmas Indicopleustes wrote it down plainly: it is accepted everywhere, from one end of the earth to the other.
Stability is a policy, not a property
Gold did not make the solidus stable. Rome had gold coinage for centuries and debased it whenever the treasury ran short; the third-century currency collapsed exactly that way. What made the solidus different was the standing decision — renewed by roughly fifty emperors of wildly varying competence — not to touch the standard. The innovation was not the coin. It was the refusal.
That refusal is the hard part, and it is the part most modern monetary projects get wrong. Any mechanism can hold a peg on a calm Tuesday. The question that matters is what happens when the issuer is under pressure and quietly loosening the standard would solve this quarter's problem. The empire answered that question correctly for seven hundred years — and then, starting around 1034, incorrectly for fifty. Small debasements, each defensible, each temporary. The coin fell from twenty-four carats toward eight, and a reputation compounded over seven centuries was spent in a generation. It never came back.
Durable money is an engineering problem
The lesson I take from both halves of that story: durability fails from discretion, so remove the discretion. Where the Byzantines relied on institutional willpower, we can rely on structure — reserves that are attested cryptographically rather than promised, rules that are enforced by mechanism rather than by memo, systems where keeping the standard is the default and breaking it is loud, visible, and hard.
That is not a metaphor; it is the actual work. I have spent twenty years on the machinery of engineered trust — payments, identity verification, zero-knowledge proofs, verifiable credentials — and the pattern is always the same: every place a promise can be replaced by a proof, replace it. The solidus did not ask to be believed. It could be weighed.
Why this is our thesis
Byzantine Solidus exists because we think the stablecoin era's real product is not any single token — it is durable digital money as infrastructure, and infrastructure is built by companies that plan in decades. Our first holding, SafeBank, is the financial operating system for that money: accounts, payments, yield, and cards for people, businesses, and AI agents, built by people who have spent careers on the rails underneath.
Seven hundred years is the benchmark. Everything we build gets measured against a coin.